Face recognition technology is a biometric-based recognition method and a technology that uses inherent physiological or behavioral features of humans which may uniquely identify their identities for authentication. The generalized face recognition technology includes a series of related technologies for constructing a face recognition system, the related technologies including face image acquisition, face localization, face recognition and preprocessing, identity confirmation and identity search. In the narrow sense, the face recognition refers specifically to a technology or system for face recognition or identity search through face. With the progress of science and technology, the face recognition technology has been greatly improved and plays an important role in many fields.
In a practical application, in the face recognition technology, an image including a face may be acquired first, then the image including the face is recognized, and finally a recognition result is acquired. The quality of the image including the face is directly related to the quality of the face recognition result. However, the lenses or cameras in the existing terminals (such as face recognition gate, face recognition attendance machine and face recognition payment machine) with face recognition functions are mostly used in indoor environments or environments with sufficient lights. These lenses or cameras have high requirements for lights, and the quality of images obtained under poor light conditions is low, which directly cause the acquired images including a large number of images that can not be used for the face recognition, thus reducing the accuracy of face recognition